Many claim that substance abuse evaluation research has failed to adequately conceptualize clinically relevant aspects of the treatment process or adequately measure treatment effects and thus failed to find that treatment has much effect. The goal of this study is to examine the relationships among patient characteristics, life context factors, treatment processes and outcomes in the treatment of drug problems using an improved evaluation design. Theory-based measures of treatment processes will be used to assess treatment effects in the context of ar expanded evaluation paradigm. The specific focus of this study is on evaluating the 12-Step chemical dependency treatment model, since processes mobilized by this approach are claimed to be curative, but have received little study. Subjects (N =400) in residential and outpatient chemical dependency treatment programs will be assessed at baseline and followed at 6 weeks, and at 6, 12 and 24 months. Illness severity, life context, treatment process and outcomes will be assessed. Structural equation modeling will be used to assess study hypotheses. The study will provide a rigorous test of the treatment theory that underlies 12-Step treatment approaches by assessing whether processes hypothesized as curative by this approach such as accepting powerlessness actually mediate outcomes. In addition, the study will examine whether processes mobilized by this approach may be more effective for some types of patients, but less effective for others. Several alternative hypotheses concerning predictors of outcome will also be considered. In addition, the study will assess whether the curative effects of 12-Step treatment are specific and unique or are able to be subsumed under a broader set of change factors. Study results should provide a clearer understanding of how change occurs in substance abuse treatment and enhance ability to match patients to treatment.